JP2000115090A - Data broadcast scheduling method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a data broadcast scheduling method which actualizes nearly ideal data broadcast wherein broadcasting cycles of information are fixed and a wait time is shortened while an arithmetic load is reduced. SOLUTION: When information to be broadcasted is all divided by a predetermined number of groups after being rearranged in the decreasing order of boadcasting rates, the information is grouped by allocating the information to the respective groups so that the sum of broadcasting rates of the information included in each group is equal among the respective groups. Then the respective pieces of information are divided by pages according to data broadcasting units and the output order of the pages in the respective group is fixed. Pages to be broadcasted are selected while selecting in order and circulating groups and in each group, selecting in the output order and circulating pages, and then the broadcasting schedule is determined to broadcast the information.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data broadcasting schedule system for a broadcasting station which enables efficient reception on the whole receiving side when a broadcasting station broadcasts a plurality of types of information.

[0002]

2. Description of the Related Art Data broadcasting is one of the most effective means for terminals to obtain information in a mobile environment. Within the radio coverage area, there is no direct relationship between the number of broadcast recipients and the required bandwidth, and the more recipients the more effective use of the bandwidth.

[0003] A data broadcasting station periodically and repeatedly broadcasts various kinds of information. The information includes stock prices, weather forecasts, traffic information, flight departure / arrival information, etc., and an unspecified number of recipients who carry a terminal within the coverage area need their own information from among multiple broadcasted information Select and receive appropriate information.
If the received information is stored in a local file of the terminal, it can be used at any time. In a mode in which data broadcast by a data broadcasting station is passively received via radio, it can be considered that the receiver has a database in the air.

[0004] In such data broadcasting, one of the items to be considered when designing a broadcasting schedule for determining the output order of data from a data broadcasting station is to include information desired by a receiver after starting reception. There is an average value of the waiting time until reception is possible. Here, the ratio of the information actually received by the receiver among all information broadcasted by data is defined as a reception rate. Assuming that there are 100 recipients and each recipient receives only one piece of information, 3
The reception rate of information received by 0 recipients is 30%. If there is a difference in the above reception rate depending on the type of information,
The data broadcasting station differentiates the broadcasting frequency of each information, and broadcasts the information with high reception rate at a large frequency (short repetition cycle),
If information with a low reception rate is broadcasted at a low frequency, it can be easily estimated that the average waiting time for the entire receiver can be shortened. For example, information having a high reception rate (many listeners) is every two minutes, and information having a low reception rate (many listeners) is 10 minutes.
Broadcast every minute. This data broadcasting frequency, that is, how to determine the distribution ratio of each information broadcast under given broadcast band conditions is a very important factor in providing efficient data broadcasting to the receiver. Becomes

According to existing studies, it is optimal to make the broadcast distribution ratio to the information to be broadcast proportional to the square root of the reception rate of each information in order to minimize the average waiting time for data reception per receiver. It has been revealed. However, even if this optimum broadcast distribution ratio can be proved theoretically, if one of the fractions appears during the calculation or the result of the calculation results in multiple pieces of information being allocated to the same transmission timing, one of them is shifted. It is not easy to achieve optimal data broadcasting, such as broadcasting. Conventionally, two methods have been proposed to realize this in an ideal state. In the following description, information to be broadcast is treated as one data chunk,
This is called a file. The basic unit of the file data size is called one page, and the time for broadcasting one page is called one time slot. Therefore, the receiver receives information for each page allocated to the time slot which is a broadcast unit. If the information is of variable length, one file is composed of one page or a plurality of pages. For a file having a plurality of pages, the reception rate of each page can be obtained by dividing the reception rate of the file by the number of pages of the file.

One of the techniques for realizing an optimum data broadcast is to select a page having the maximum value obtained by multiplying the time from the last broadcast of each file or page to the present by the output distribution ratio. This calculation can be simplified by adding a period-related penalty only to the selected file or page. This method is called the periodic individual selection method (I
DS method: Individual Selecting Method).

[0007] Another is NHVaidya et al., "Scheduling D
ata Broadcast in AsymmetricCommunication ", Proc.of
Workshop on Satellite-based Information Services (W
OSBIS), a method called the bucket method proposed in New York November 1996. In this method, processing is performed not on a page basis but on a file basis. Broadcast files are divided into several groups, and each group is sequentially broadcasted to narrow down the selection of output file selection. The method of grouping is as follows.

[0008] The number of divided groups is numgrp,
The total number of pages of all files to be broadcast is represented by m, and the number of each page is represented by i. At this time, i = 1 to m.
The broadcast ratio of page i to the entire band is w _i , w ₁
≧ w ₂ ≧ ... ≧ w _m, and def = (w ₁ −w _m ) / numg
rp. Also, w _i is w ₁ − (k−1) * def ≧
w _i > w ₁ -k * def (k = 0 to an integer of numgrp)
It belongs to the group k that satisfies It is assumed that w _m belongs to the group k (= numgrp). Let pb be the average reception rate of each group. Note that the broadcast ratio refers to a ratio of each information broadcast to all information. In this case, selected one by one file of the candidate from each group, pb × t _i ² (waiting time t _i is page i) selects the largest file. The selected file is put at the end as the next output candidate in the group.
By employing this method, the calculation for selecting an output candidate can be simplified.

[0009]

However, the conventional IDS method can realize an almost ideal data broadcast with a smaller average waiting time than the bucket method, but the amount of calculation is enormous. In addition, the broadcasting cycle of certain information cannot be made constant.

Although the bucket method is easier to calculate than the IDS method, it still requires a lot of calculations. Also, I
Not as ideal as the DS method. In addition, the broadcasting cycle of certain information cannot be made constant.

The fact that the broadcast cycle of one piece of information cannot be made constant causes the following problem. For example, if there is information that is theoretically optimal to be broadcast every 10 minutes, when a fraction is calculated by calculation, 9.9 minutes after the immediately preceding broadcast is performed, and at some point 10.1
Or broadcast in minutes. As described above, if there is a possibility that the broadcast cycle is shifted, the receiver must receive the broadcast earlier and longer with a margin in order to reliably receive the information. Although any receiver desires efficient reception, especially when using a terminal powered by a rechargeable battery in a mobile environment,
In particular, it is desired to broadcast data at a predetermined time and at a predetermined cycle from the viewpoint of saving battery.

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and an object of the present invention is to make it possible to fix the broadcast cycle of information, reduce the computation load, and reduce the waiting time. An object of the present invention is to provide a data broadcasting schedule method for realizing near data broadcasting.

[0013]

In order to achieve the above object, a data broadcasting schedule method according to the present invention occupies all information in a data broadcasting station which periodically and repeatedly broadcasts a plurality of types of information. Arranging the information in descending order of the broadcast ratio representing the broadcast ratio of each information;
When dividing all information into a predetermined number of groups, by allocating information arranged in order from the highest broadcast ratio from one group so that the sum of the broadcast ratios of the information included in each group becomes equal. A grouping step of forming a group, and a broadcasting step of selecting each group in order and broadcasting the selected information while sequentially selecting information in the selected group. The information to be broadcast is determined while circulating.

In the above invention, when the information is larger than a data broadcast unit, the information is divided into pages corresponding to the data broadcast unit, and the information to be broadcast is divided into groups by pages and within groups. It broadcasts while circulating.

[0015]

Preferred embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a schematic overall configuration diagram of a data broadcasting system employing a data broadcasting scheduling method according to the present invention. The data broadcasting station 1 periodically and repeatedly broadcasts various types of information. A broadcast receiving terminal (hereinafter, simply referred to as a “terminal”) 2 within the coverage area of the data broadcast station 1 receives only information necessary for itself from information broadcast by the data broadcast station 1.

FIG. 2 is a block diagram of a server installed in data broadcasting station 1 in the present embodiment. The server 10 includes: an information storage unit 11 for storing information to be broadcast;
It is composed of a broadcast schedule determining unit 12 that determines what broadcast schedule the information should be broadcast on, and a data broadcast unit 13 that broadcasts information according to the determined broadcast schedule. The server 10 according to the present embodiment can be realized by a computer having a general wireless broadcast function, and has a feature in processing in the broadcast schedule determination unit 12.

A feature of the present embodiment is that all information is divided into groups so that the sum of the broadcast ratios of the information allocated to each group is equal, and the output order of the information in each group is determined. It is immobilized. When the output order of the information broadcast from the data broadcasting station 1 is determined, each group is selected in turn, and the information is selected in order within the group in accordance with the fixed information output order. That is, selection is performed while circulating between groups, such as the first page of group 1, the first page of group 2, the first page of group 3, and the second page of group 1. As described above, by mechanically selecting information to be broadcast while circulating between groups and circulating within the group, the information broadcasting cycle can be fixed, and the computational load can be reduced. While reducing the average waiting time to a level close to ideal. In this embodiment, since a double circulation is formed between groups and within a group, the double circulation method (D
CS method: Double Cyclic Sequencing Method).

Next, a procedure for determining a data broadcasting schedule and performing data broadcasting according to the present embodiment will be described with reference to FIG.
This will be described with reference to the flowchart shown in FIG. This process is performed by the broadcast schedule determining unit 12. In the present embodiment, all information is divided into a plurality of groups as described later, and the number of divided groups (numgrp)
Must be determined in advance. In the present embodiment, the data is divided into three groups. It is also assumed that the broadcast ratio of each information is known. Each piece of information is assumed to be formed of one file and has a variable length.

First, all information to be broadcast is arranged in ascending order of broadcast ratio (step 100). That is, the total number of information is m,
Assuming that the broadcast ratio of each information is w _i (i = ₁ to m), w ₁ ≧
Arrange so that w ₂ ≧... ≧ w _m . For example, the number of information m =
Assuming that information A, B, C,..., And K are arranged as a result of using the 11 pieces of information A to K in descending order of broadcast ratio, the broadcast ratio of information A is w ₁ and the broadcast ratio of information B is is so on w _2, can be expressed as _{w 1 ≧ w 2 ≧ ... ≧} w 11. After arranging information in this manner, information is grouped by a method which is a feature of the present embodiment (step 200). The details of this information grouping process will be described with reference to the flowchart shown in FIG.

The present embodiment is characterized in that all information to be broadcast is grouped so that the sum of the broadcast ratios of the information included in each group is equal. First, assuming that the group number is grp, first, grp =
1, that is, a group 1 is formed (step 201). This adds the broadcast ratio of each information in order from the information A with the highest broadcast ratio. The added broadcast ratio is 1/3 (= grp / n) of the broadcast ratio (100%) of all information.
umgrp) (step 20)
2). When grouping of group 1 is completed, group 2 is formed in the same manner as above (step 20).
3, 202). That is, the process is repeated until the sum of the broadcast ratios of the information reaches 2/3 (= grp / numgrp) of the broadcast ratio of all information. This is called group 3
(Step 204).

FIG. 5 is a diagram showing the files (information) A to K arranged in descending order of the broadcast ratio, and further showing their page lengths and groups. In the present embodiment, the information A is stored in the group 1 as shown in FIG.
The information BE is divided into group 2 and the information FK is divided into group 3. This group 1,
The arrangement of 2 and 3 is the selection order when circulating the group.

In the present embodiment, information is grouped as described above. The basic idea of this grouping process is based on the following conditions.

FIG. 6 is a diagram showing the relationship between files (information) and broadcast ratio when files are arranged in descending order of broadcast ratio. Grouping all information to perform ideal data broadcasting is also performed by the conventional bucket method. However, in the bucket method, the maximum value and the minimum value of the broadcast ratio are obtained, and the interval is divided equally by the number of divided groups to perform grouping. On the other hand, in the present embodiment, in FIG. 6, the area of the area surrounded by the coordinate axes, the number of files, and the value of the broadcast ratio is equally divided by the number of division groups. The areas A1, A2, and A3 of the hatched portions divided into the number of divided groups (3 in this example) correspond to the sum of the broadcast ratios in each group, and in the above grouping process, A1 = A2 = A3. That is, the processing is performed.

When the information can be grouped as described above, the data to be broadcast is determined as follows (step 300). This processing will be described with reference to the flowchart shown in FIG.

First, since data broadcasting is performed by allocating one page to one time slot, if each information has a fixed length of one page, the data may be left as it is. However, fixed length data formed from a plurality of pages or variable length data may be used. In the case of, the corresponding information must be divided into pages. Since each of the information AK broadcasted in the present embodiment is formed from pages 1 to 5 as shown in FIG. 5, information other than the information C and H having one page length is divided into page units ( Step 301). The arrangement of each piece of information after division is maintained. The page arrangement at this time is as shown in FIG. 8A, and this arrangement order is the output order in each group. This output order is fixed. When the arrangement of pages is determined in this way, the broadcast schedule determination unit 12 selects pages in the following order and causes the data broadcast unit 13 to broadcast the pages.

First, assuming that the group number is grp and the arrangement order of the pages in each group is i _grp , each is initialized with 1 representing the head (step 302). First,
grp = 1, that is, the first (i ₁ = 1) page of group 1 containing the information with the highest broadcast ratio is selected and broadcast (step 303). Then, the next group 2
Is selected and broadcast (step 304,
309, 307, 303). Furthermore, the next group 3
Is selected and broadcast (step 304,
309, 307, 303). The group number (grp) after this broadcast is the number of divided groups (numgrp)
3 (step 304). That is, since the group 3 is the last group, the process returns to the first group 1 (step 305), and after selecting the second page following the first page of the group 1 (step 306),
It is broadcast (steps 307 and 303).

The processing up to this point will be described using a specific example. Based on the above-described processing, the first page “A1” of group 1, the first page “B1” of group 2,
The first page of group 3 is broadcasted as “F1”. When the group 3 is selected, the process returns to the group 1. However, since the first page “A1” is selected in the immediately preceding broadcast in the group 1, the second page “A1” following this page is selected.
This data broadcasting process is repeated, and when the fourth page “F4” of group 3 is broadcast, the program returns to group 1, but the fourth page “A4” is broadcasted earlier. There is no page following this.

Returning to FIG. 7, since there is no page following the last (fourth) page broadcast immediately before returning to group 1 (step 307), the first page is selected (step 307). Step 308) and broadcast it (Step 303). Then, in the next group 2, there is a fifth page following the fourth page broadcast immediately before, so that page is selected and broadcast (steps 304, 309, 307, 303).

This process will be described with a specific example. When the fourth page "F4" of group 3 is broadcast, the program returns to group 1, but in the immediately preceding broadcast of group 1, the last page "A4" is displayed. Since the broadcast has been made, the process returns to the first page “A1”. That is, broadcasting is performed while circulating pages within the group. On the other hand, in group 2, since the fifth page "D2" exists after the fourth page "D2", the page "D2" is broadcast without returning to the first page "B1".

According to the above-described processing, the first page "B1" is broadcast after the eleventh page "E4" is broadcast in the group 2 as well. After broadcasting the eighteenth page “K2” in group 3 as well, the first page “F1” will be broadcast.

The broadcast schedule is determined as described above. The order of the broadcast pages is shown in FIG. As is apparent from this output order, the broadcast frequency of the information A belonging to the group 1 is the highest, and subsequently, the information B to E belonging to the group 2 has a lower broadcast frequency than the group 1. Information F to K belonging to group 3 has a lower broadcast frequency than group 2. In the example shown in FIG.
The page “A1” is broadcast three times before the page “E4” of the group 2 is broadcast and five times before the page “K2” of the group 3 is broadcast. This is natural because the information is arranged in descending order of the broadcast ratio.

In the present embodiment, as described above, broadcasting is performed in page units while circulating between groups and circulating within a group. In the conventional bucket method, after dividing into groups based on FIG. 6, calculation from which group information is to be extracted and broadcast was performed, so that the calculation load was heavy. After the division into groups, only the circulation is performed mechanically, so that the calculation for determining the page to be broadcast is unnecessary.

Based on the broadcast schedule in this embodiment in which the output order between groups and within each group is fixed, it is possible to know in advance when each page will be broadcast. In other words, since the broadcast cycle of each page assigned to a time slot is fixed, if the terminal receiving the broadcast knows this cycle in advance, it is necessary to start up and operate the system only during the broadcast time. Thus, desired information can be reliably received. For example, if the information is broadcast every five minutes, the mobile computer may automatically start up every five minutes, receive data, store the data, and perform an automatic shutdown. In this embodiment, since the broadcast is reliably performed every 5 minutes, it is not necessary to start up the system and wait for an extra time. Further, the information is divided into groups so that the sum of the broadcast ratios in each group becomes equal, and the information is circulated between the groups and within the group, so that the average waiting time can be shortened.

The DCS method described in the present embodiment means that when a file is composed of a plurality of pages, each page of one file is not continuously output but is output with a gap. (Hereinafter referred to as “alternate method”)
). Assuming that the number of groups in the cycle is nmc and the variable length file is nump pages,
numc × (nump-1) +1 unit time is required,
Compared to the case of nump unit time for continuous reception, (numc
-1) × (nump-1) extra time is required for the unit time. However, instead, the cycle is stabilized for each page of each file. That is, in the bucket method, the variable length file is treated as one output unit, and is not divided on the way (hereinafter, referred to as “serial method”).

However, the alternating method does not necessarily work in all cases at a disadvantage over the serial method. Here is a simple example. Assume that the waiting time is to be evaluated and that there are two types of files (information) A and B. A is 4 pages long and B is 1 page long. The reception frequency of B is four times that of A. The output ratio of A and B is (１ ／) from the theory that α = 0.5 is optimal.
^1/2 : (4/1) ^1/2 = 1: 4 is preferred. At this time, the serial system and the alternating system are compared as shown in FIG.
In the case of the serial method (bucket method), the wait of A is 7 on average,
The wait for B averages 1.75. Alternating method (DCS method)
In this case, the wait of A is 11 on average, and the wait of B is 1 on average. When the weighted average is calculated based on the reception rate, the serial method is 2.8 and the alternate method is 3.0 in the wait, and the serial method is advantageous. However, similarly, when calculating the case where A is 8 pages long, B is 1 page length, and the reception frequency of B is 8 times A, the wait of the serial system is 3.7 and the wait of the alternate system is 3. 4
And the alternating method is advantageous. Thus, the stability of the cycle also greatly affects the waiting time. That is, according to the example shown in FIG. 8, it is considered that a longer waiting time occurs after the page “A1” constituting the information A is received until the last page “A4” is broadcast. The average waiting time of the entire data broadcast can be shortened, and the broadcast cycle can be stabilized. In addition, the distortion of the waiting time (difference in length) can be reduced.

Next, the effectiveness of the DCS method described in this embodiment will be evaluated by comparing it with a conventional method. For comparison, the FLAT method that outputs all files and pages equally, the IDS method that is an ideal method for minimizing the waiting time and delay time, and the bucket method that has been proposed in previous research, are used. select. As an evaluation index, we used the waiting time mentioned in the previous study to make the comparison easier to understand. The average waiting time in the broadcast schedule determined by each method is confirmed by probability calculation or simulation.

The file length is a fixed length type in which all files are one page, and the length of each file is one.
Two patterns of indefinite length given by uniform random numbers of 10 pages to 10 pages were prepared. As other parameters, the number of broadcast files = numdata, the number of groups of files adopted by the bucket method and the IDS method = numgrp, the power parameter for obtaining the delay time coefficient from the file distribution density = α, and the density of update files are used. I do. FIG. 10 shows the conditions of this parameter.

FIGS. 11 and 12 show the case where file length = fixed length, numgrp = 5, α = 0.5, and numdata
a = 100,1000, FLAT method, ID
S method, bucket method, DCS method, and θ = 0
It is the result calculated about the combination of Zipf distribution when it is set to 1.5. The following can be said as a result of the calculation.
(1) The greater the distortion of the Zipf distribution, the shorter the average waiting time. (2) The IDS method is the most effective, but the DCS method is generally more effective than the bucket method, and there is no significant difference from the IDS method. (3)
When the number of files is large, the difference between the DCS method and the IDS method is small, whereas the difference between the bucket method and the IDS method is large.
This is particularly noticeable when the distortion of the distribution is large.

FIG. 13 shows the relationship between the number of groups and the waiting time in the bucket method and the DCS method. α = 0.
5, Zipf distribution coefficient θ = 1.5, number of files number
Assuming that data = 1000, values of numgrp = 2 to 20 are taken.

As the number of groups increases, the waiting time gradually decreases. However, the DCS method reaches saturation when the number of groups is increased, whereas the bucket method continues to decrease gradually.
This is because the DCS method mechanically circulates the allocation. Even if the number of groups is excessively increased, the imbalance in the distribution among the groups increases, and approaches the FLAT method. Is performed each time, and if the number of groups is increased, the IDS method is approached. In the above example, it can be said that at most 6 to 7 groups are sufficient with the DCS method. In addition, it goes without saying that increasing the number of groups adversely affects the calculation load.

Calculation is performed for a case where the file length is variable. FIG. 14 shows distributions of reception rates, numdata = 100, and FLA when numgrp = 5 and α = 0.5.
It shows four methods of T method, IDS method, bucket method, and DCS method. IDS as in the case of fixed length
Law waiting time is the shortest, DCS method, bucke
The t method and the FLAT method are in order.

FIG. 15 shows the case where the file length is indefinite and the reception rate distribution is θ = 1.0, numdata = 100, numg
RP = 5, the number of pages is 1, 10, and 100, respectively, showing three methods of the IDS method, the bucket method, and the DCS method. Even if the number of pages is large, the DCS method is generally more efficient than the bucket method, but the deviation from the ideal value is large. Since the pages in the file are output only intermittently, the waiting time is relatively disadvantageous. If you determine the page unit in detail, the loss on the band will be less,
Since the number of pages increases and the waiting time becomes disadvantageous, there is a trade-off between the bandwidth efficiency and the waiting time.

[0044]

According to the present invention, data is broadcast while mechanically circulating the output order between and within groups, so that the arithmetic processing when determining the broadcast schedule can be greatly reduced. It is possible to fix the broadcast cycle of information that is repeatedly broadcast.

When information is grouped, the information is divided into groups so that the sum of the broadcast ratios of the groups becomes equal.
Since the information to be broadcasted is selected from each group in order, the average waiting time until the entire broadcast data can be received can be reduced.

Although the method of the present invention does not achieve the theoretically desired ideal, it is possible to realize a data broadcast close to the ideal with a simpler calculation than the conventional IDS method or bucket method.

[Brief description of the drawings]

FIG. 1 is a schematic overall configuration diagram of a data broadcast system employing a data broadcast schedule method according to the present invention.

FIG. 2 is a block diagram of a server installed in a data broadcasting station according to the present embodiment.

FIG. 3 is a flowchart showing a procedure for performing a data broadcast by determining a data broadcast schedule in the present embodiment.

FIG. 4 is a flowchart illustrating information grouping processing according to the present embodiment.

FIG. 5 is a diagram showing an arrangement order of information to be broadcast, a page length, and a group to which the information belongs in the present embodiment.

FIG. 6 is a diagram illustrating a relationship between information and a broadcast ratio when files are arranged in descending order of the broadcast ratio.

FIG. 7 is a flowchart showing a procedure for determining a broadcast schedule in the present embodiment.

FIG. 8 is a diagram showing pages to be broadcast and an output order thereof in the present embodiment.

FIG. 9: Information (page) in the serial system and the alternating system
FIG. 3 is a diagram showing an example of arrangement.

FIG. 10 is a diagram showing, in a table form, the contents of parameters defined when evaluating the DCS method in the present embodiment.

FIG. 11 is a diagram showing a waiting time in each method when a file has a fixed length.

FIG. 12 is a diagram illustrating a waiting time in each method when a file has a fixed length.

FIG. 13 is a diagram showing the relationship between the number of groups and the waiting time in the bucket method and the DCS method.

FIG. 14 is a diagram showing a waiting time in each method when a file has a variable length.

FIG. 15 is a diagram showing a waiting time in each method when a file is of an undefined length type.

[Explanation of symbols]

1 data broadcasting station, 2 broadcast receiving terminal (terminal), 10
Server, 11 information storage unit, 12 broadcast schedule determination unit, 13 data broadcast unit.

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Masahiro Aono 2-3-2 Marunouchi, Chiyoda-ku, Tokyo Within Mitsubishi Electric Co., Ltd. (72) Inventor Takashi Watanabe 347-1 Arita Minamicho, Hamamatsu City, Shizuoka Prefecture Art Pia Aritama Minami 801 (72) Inventor Tadanori Mizuno 175-1 Tomizuka-cho, Hamamatsu-shi, Shizuoka

Claims (2)

[Claims] 1. A data broadcasting station which periodically and repeatedly broadcasts a plurality of types of information, arranging the information in ascending order of a broadcast ratio representing a ratio of each information to all information broadcasted; Groups that form groups by allocating information arranged in descending order of broadcast ratio from one group so that the sum of the broadcast ratios of the information included in each group is equal when dividing into groups A broadcasting step of selecting each group in order, and broadcasting the selected information while sequentially selecting information in the selected group, and broadcasting while circulating between and within the group. A data broadcasting schedule method, which determines information to be performed. 2. The method according to claim 1, wherein, if the information is larger than a data broadcast unit, the information is divided into pages corresponding to the data broadcast unit, and information to be broadcast is circulated between and within groups in page units. 2. The data broadcasting schedule method according to claim 1, wherein broadcasting is performed.